That's a good point—I wonder if temperature during application could also play a role? I've noticed sealants applied in colder weather seem to fail so...